Rate Splitting for Uplink NOMA with Enhanced Fairness and Outage Performance

Hongwu Liu, Theodoros A. Tsiftsis, Kyeong Jin Kim, Kyung Sup Kwak, H. Vincent Poor

Research output: Contribution to journalArticlepeer-review

61 Scopus citations


In this paper, we investigate rate splitting (RS) for an uplink non-orthogonal multiple access (NOMA) system with a pair of near and far users adopting cyclic prefixed single carrier transmissions. Frequency-domain equalization is applied to assist successive interference cancellation at the base-station. Two kinds of RS schemes, namely, fixed RS (FRS) and cognitive RS (CRS) schemes, are proposed to realize RS for uplink NOMA with the aim of improving user fairness and outage performance in delay-limited transmissions. Corresponding to the split data streams, transmit power is allocated in either a fixed or cognitive manner for the FRS and CRS schemes, respectively. Based on achievable rate region analysis, the benefits of applying RS to uplink NOMA for enhancing the user fairness and outage performance are revealed. A modified Jain's index is proposed to measure the user fairness for the considered delay-limited transmissions. Closed-form expressions are derived for the outage probabilities of the paired users, respectively, whereas the preferred system parameters are chosen based on asymptotic outage probability expressions. The enhanced user fairness and superior outage performance of the proposed RS schemes are corroborated by Monte Carlo simulation results.

Original languageEnglish (US)
Article number9064705
Pages (from-to)4657-4670
Number of pages14
JournalIEEE Transactions on Wireless Communications
Issue number7
StatePublished - Jul 2020
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Applied Mathematics


  • NOMA
  • outage probability
  • rate splitting (RS)
  • user fairness


Dive into the research topics of 'Rate Splitting for Uplink NOMA with Enhanced Fairness and Outage Performance'. Together they form a unique fingerprint.

Cite this